Risk Assessment for Follicular Lymphoma

Loretta J. Nastoupil, MD:For newly diagnosed patients with follicular lymphoma, we’re relying on our risk prognostic tools, which are primarily utilizing clinical criteria. The criteria that we often apply is the FLIPI score, which incorporates age, stage, LDH. The extent of lymphoma involvement, meaning the number of lymph node regions that are involved, it will help us when we’re comparing across studies because most of the low or intermediate FLIPI patients do quite well. It’s the high-risk patients that are still associated with inferior outcomes. But in terms of choosing therapy, the FLIPI score has its limitations. It helps us when we’re discussing with patients their prognosis, which is sometimes useful for a given patient. Sometimes it can also confuse patients. But I think it’s also helpful with us when we’re risk-stratifying patients, meaning how worried do we need to be about a patient who’s high-risk FLIPI versus lower intermediate—which, again, we believe those patients to do quite well.

The limitation of the FLIPI score is that it has not incorporated molecular features to help us choose one therapy over another. If I have a patient with a high-risk FLIPI score, I do not know if they’ll do better with chemotherapy/rituximab or if they’ll do fine with an immune therapy-based approach. So, there is a current unmet need in follicular lymphoma where we need better risk-stratifying tools. There has been a recent advancement in terms of the M7FLIPI where there was an approach to incorporate molecular findings into the risk score. It was a retrospective series that was validated where they examined the molecular features in addition to the FLIPI components. And what we learned is that it was valuable, but the FLIPI score and the ECOG performance status still were prognostic in terms of how well or poorly patients did.

We also learned that there was a subset of patients with poor-risk FLIPI who actually were going to do quite well, and that was primarily driven byEZH2mutations. So, there is more to learn in terms of the biology of this disease. And the M7FLIPI, though it is an advancement, it’s not clinically applicable, meaning you can’t derive those molecular features from a current path report.. But at least it is a step forward in trying to incorporate molecular findings into our risk stratification or prognostic tools. M7FLIPI right now is currently something that is confined to research studies. Again, it’s very useful identifying the patients who are going to do poorly with chemoimmunotherapy, but it is not something that the average clinician can apply. There needs to be tools that are applicable in the clinic if they’re going to be adopted, and if they’re going to inform practice decisions, and we’re not there yet with the M7FLIPI.

With the median age of diagnosis for patients with follicular lymphoma around 60 to 62 years of age, oftentimes these patients will have comorbid conditions. In contrast to our CLL colleagues, we’re not routinely incorporating comorbidity scoring into our decision making. For instance, the CIRS score, which is often utilized to decide if a patient is a candidate for chemotherapy or not in CLL, we’re not currently adopting or applying those types of scoring systems into deciding is this patient appropriate for bendamustine plus CD20 antibody versus rituximab monotherapy. I think that is currently an unmet need in follicular lymphoma, whether we want to adopt validated scoring systems into our prospective clinical trials to distinguish appropriate candidates for intensive therapy versus less intensive therapy. That needs to be explored.

So, to summarize this case, we have a 71-year-old patient with limited comorbidities, with the exception of underlying rheumatoid arthritis, who has been on immune suppressive therapy in the form of methotrexate and who presents with asymptomatic bilateral axillary swelling. On confirmation of a diagnosis of follicular lymphoma, staging has revealed that this patient is at least stage 3, but is of low tumor burden as assessed by the GELF criteria. This patient is an appropriate candidate for observation. One may question the methotrexate use and whether or not there’s an EBB-associated lymphoma component to this or whether or not treating the lymphoma will have implications for the underlying autoimmune condition. But I still think that it’s appropriate to apply the GELF criteria when deciding when a patient is an appropriate candidate for therapy.

In regards to the discussion I often have with patients who have a new diagnosis of follicular lymphoma, if they meet the GELF criteria, the alternative—meaning if they have low tumor burden, what would be the alternative approach—I will oftentimes apply is the RESORT criteria, where they took patients who were all of low tumor burden and treated them with 4 weekly doses of rituximab and then looked at a follow-on approach—whether it was retreatment or maintenance—and we saw that there was no difference between those 2 arms. So, if I have a low tumor burden follicular lymphoma patient who is uncomfortable with the idea of observation or there is an indication for treatment, but not quite the GELF criteria, I will offer 4 weekly doses of rituximab followed by observation. Oftentimes when I recommend observation or watchful waiting to a patient, there is some resistance because they have been rendered a diagnosis of cancer and they’re uncomfortable with the idea of doing nothing proactive. I think there’s a lot to be gained by explaining why we observe patients. Again, there is that small opportunity that they may have spontaneous regression, and I’ve seen it, particularly in patients with low tumor burden, meaning low-volume lymph nodes. So, that’s one benefit to observation.

One benefit of observation is that I can guarantee they will have no side effects with therapy, since we are observing them. Any treatment, even if it’s rituximab monotherapy, will carry some risk, and each patient’s reaction to a given therapy can be highly variable. Secondly, I let them know that we are not just watching and waiting until they get into trouble, but we’re monitoring until there is an objective finding that suggests they’re no longer appropriate for observation. And we’ve learned through retrospective studies that this will not render an increased risk to that patient. And it’s a careful conversation to have, but I think that it can be done in a way where patients understand that observation can be appropriate for them. Despite that, I still have patients that want a plan to include something proactive and they’re less concerned about the side effects of that therapy or how I choose one therapy over another. Again, revisiting this patient who is of low tumor burden by the GELF, I think rituximab monotherapy is a reasonable option based on a prospective study, and I will follow the treatment scheme of the RESORT study while I will administer 4 weekly doses of rituximab and then observe those patients.

Transcript edited for clarity.

January 2014

• A 71-year-old female reports having symptoms of bilateral axillary swelling of 1.5 years’ duration and presents with diffuse inguinal and cervical adenopathy.
• Past medical History: 15-year history of treatment for rheumatoid arthritis with methotrexate
• Physical examination:
  • The patient is generally well-appearing; temperature, pulse, blood pressure, and HEENT are all WNL; extremities show no edema.
  • Cardiac exam is normal; chest is clear
  • Abdomen shows no abnormal hepatosplenomegaly
  • Lymph nodes: left axillary 1.5 cm, right axillary 2 cm; cervical and inguinal nodes <1 cm bilaterally; non-tender
• Notable laboratory findings:
  • CBC with diff, WNL
  • LDH, 148
• Right groin excisional node biopsy shows small lymphocytes with nuclear indentations (centrocytes) and large lymphocytes without indentations (centroblasts).
  • Pathology: t(14;18); co-expression of Bcl2, CD10, CD20.
• CT shows scattered adenopathy in the cervical, axillary, mesenteric, and pelvic regions. The largest lymph node measures 4.5 cm. The remaining lymph nodes are smaller than 3 cm.